本文给出了在有限的频率范围内近似计算离散数据序列能量谱密度(ESD)的一种新方法。该方法利用了关于某一频率展开的泰勒级数的截尾形式,它比 FFT 内插速度更快.利用余数定理,可以用解折方法证明近似误差是有界的.作为应用的一个例子,该方法用来估计中心频率。 This paper presents a new approach to approximate the energy spectral density (ESD) of discrete data sequences over a limited frequency range. The method uses a truncated version of the Taylor series expanded at a certain frequency, which is faster than the FFT interpolation.Using the remainder theorem, the unfolding method can be used to prove that the approximation error is bounded.As an example of the application, This method is used to estimate the center frequency.